From real-world views to geographical information

Position paper for workshop on Fundamental Issues in Spatial and Geographic Ontologies, Ittingen, 23rd September 2003

Held in conjunction with COSIT 2003

Jenny Harding, Research & Innovation, Ordnance Survey

Jenny.Harding@ordnancesurvey.co.uk

The subject of this paper reflects long-term research interests of Ordnance Survey, and not necessarily the focus of current Ordnance Survey business.

The need for research

Success of geographical information providers, such as Ordnance Survey, depends on many factors within the industry environment in which they operate. Fundamental to their success is the usefulness and value of information content provided to the end user. As the national mapping agency for Great Britain, Ordnance Survey supplies geographic information products to a wide spectrum of end users ranging, for example, from central government policy makers to the person in the street. Each has their own view of the ‘real world’ depending on what they are doing, what problems they are trying to solve, what decisions they are trying to make. Within UK central government alone there are currently around 150 departments and agencies using Ordnance Survey products in different contexts for different purposes. This is a great success, but how well does current geographical information actually relate to the ‘real world’ views of so many different users?

From the end-user perspective a prime factor in determining the usefulness and value of an information source is its fitness for their particular purpose, in terms of content and quality. Other factors including pricing, licensing, supply formats and interoperability will also influence usability.

User task-focused research at Ordnance Survey

In order to understand how geographical information of the future may be more meaningfully specified and modelled to better meet the need of existing end-users and potential end-users of the future, it is necessary to better understand what matters to the user in the context of their tasks. The user’s conceptual view of the ‘real world’ in relation to their task needs be understood and described or specified, building a user-task oriented ontology.

A key research project within Ordnance Survey Research & Innovation group, is focused on analysis of tasks and decision making in which geographical things play a part, with the aim of identifying what aspects of geographical information are important or critical in the decision/task context. The research aims to identify, for example, the semantics used to identify and describe geographical things, the importance of discreteness or vagueness of conceptualised boundaries, the importance of temporal information, the significant relationships between geographical things, the level of detail that is critical to the decision/task.

Within the mass of human decision processes with a spatial element, broad differentiation can be made between those dealt with in a working or ‘professional’ capacity and those facing the general public in everyday life. As a starting point, the research is currently focusing on specific tasks within the central government area of neighbourhood renewal, an example of a 'professional' problem area requiring diverse aspects of geographical information. In contrast, we are also researching pedestrian wayfinding – a task of everyday life.

In both cases the basic approach is to first understand the problem and decision processes 'through the eyes' of those involved, then to determine how the nature and quality of geographical information features consciously or subconsciously in the process. Output from this research will contribute to understanding essential geographical information concepts and how geographical information may be meaningfully specified for use in many different user contexts.

Based on the research outcomes from these two contrasting areas, we intend to extend the spatial cognition research focus to other user task contexts. In this way a picture can be built up of different people’s conceptualisations of essentially the same geographical reality, providing input to further research tackling how to flexibly model information about geographical reality to meet the needs of diverse end-users.